ARTICLE pubs.acs.org/est
Evaluating Sources of PAHs in Urban Streams Based on Land Use and Biomonitors Sofia Augusto,† Carla Gonzalez,†,‡ Rute Vieira,† Cristina Maguas,† and Cristina Branquinho*,† †
Universidade de Lisboa, Faculdade de Ci^encias, Centro de Biologia Ambiental (CBA), FCUL, Campo Grande, Bloco C2, Piso 5, 1749-016 Lisboa, Portugal ‡ CENSE - Center for Environmental and Sustainability Research, Ecological Economics and Environmental Management Group, Faculdade de Ci^encias e Tecnologia, Universidade Nova de Lisboa, 2829-516 Caparica, Portugal ABSTRACT: Toxic polycyclic aromatic hydrocarbons (PAHs) can be found in wastewaters and sewages released from industries and/or urban areas. When discharged untreated to stream waters, they can be a problem to human health. This work represents the first attempt to use PAH and metal concentrations in aquatic moss transplants together with landuse information to identify water pollution sources in urban areas. To do this, the moss Fontinalis antipyretica was collected from a natural stream and transplanted to four different streams in a densely populated area of Lisbon, Portugal. After three months of exposure, mosses were collected and analyzed for metals and for the 16 priority PAHs recommended by the U.S. EPA. Urban streams seem to have a scattered contamination of 6-ring PAHs. Correlations among land-use, metal concentrations, and PAH concentrations indicated that areas occupied by activities of tertiary and industrial sectors had higher PAH concentrations in transplanted mosses, mainly for the sum of the 16 EPA-PAHs and for the 2-, 3- and 5-ringed PAHs, than areas occupied by urban and wooded areas. These PAHs were associated with enhanced Zn and Cu and land use activities that linked the sites to high traffic density. Industrial land use influences PAH concentration in water up to 1000 m of distance from the stream, whereas tertiary sector land use influences it up to 500 m.
1. INTRODUCTION Polycyclic aromatic hydrocarbons (PAHs) include toxic organic compounds that can be found in wastewaters and sewages released from industrial and/or urban areas; these wastewaters and sewages are sometimes discharged untreated to stream waters and can be a problem to human health. Their presence in receiving waters is associated with toxicological effects, producing endocrine disruption in marine organisms,1 neurotoxicity,2 and alterations at the ecosystem level.3 Directive 2008/105/EC specifies that hazardous PAHs should be monitored in surface waters.4 Though PAHs occur naturally in the environment, generated by forest fires and volcanic eruptions, the largest amount of PAHs is released into the environment by human activities.5 Anthropogenic PAHs result mainly from pyrolytic processes, especially the incomplete combustion of organic materials during industrial activities, home heating, power generation, incineration, and vehicle emissions, and as well as from petroleum cracking and refining in petrochemical industries, and during chemical manufacturing.5 Illegal discharges into streams are difficult to detect. Stream water concentrations only reflect very recent discharges and give little information about pollutants that may have passed through r 2011 American Chemical Society
and accumulated in biota. For this reason, the use of biomonitors can be advantageous as they can accumulate pollutants over time, revealing chronic pollutant exposure. Some of the more important toxic effects of organic pollutants like PAHs, both on biota and human health, are a result of chronic exposure at very low concentrations. Water biomonitors have been used worldwide to assess pollutants from a variety of pollution sources (metals, organic compounds, etc.). Aquatic bryophytes, specifically mosses, have proven to be effective biomonitors as they have a wide geographical and ecological distribution, lack seasonality, and are tolerant to various types of mineral and organic pollutants.6-8 However, in some areas, particularly in more disturbed ones such as urban streams, in situ aquatic mosses cannot be found. The use of moss transplants extends the range of sites to those without natural moss populations.8-11 Moreover, moss transplants allow
Received: October 27, 2010 Accepted: February 17, 2011 Revised: February 14, 2011 Published: March 16, 2011 3731
dx.doi.org/10.1021/es1036332 | Environ. Sci. Technol. 2011, 45, 3731–3738
Environmental Science & Technology the calculation of enrichment factors (EFs) from natural or less contaminated sites. Regarding specific PAHs in mosses, some studies have used transplant techniques to evaluate moss responses following exposure to these pollutants in city harbors, where levels of PAHs tend to be high and less likely subject to dilution effects.8,11 These studies have shown a strong correlation between PAH concentrations in water and PAH concentrations in moss tissues from the same sites, meaning that water pollution can be monitored using moss transplants.8 In addition, the same studies have also shown a correlation between PAH concentrations in water and adverse effects on moss physiology.8,11 One of the most used methods to assess the physiological status of mosses is chlorophyll fluorescence, which has becoming a valuable, nondestructive procedure to measure changes associated with photosystem II (PSII) due to gaseous pollutants and heavy metals.12 Though these studies reveal the potential of moss transplants to assess water pollution by PAHs, it is our understanding that no studies have used this technique to identify different pollution sources; and nowadays knowledge of the contribution of each pollutant source to stream PAH levels is needed for pollution abatement and management. To confirm the origin of PAHs, different chemical compounds, such as heavy metals, can be used as surrogates of specific pollution sources. Thus, the main objective of this work was to identify the major PAH pollutant sources in urban streams using transplanted aquatic mosses and information on land use surrounding each stream.
2. EXPERIMENTAL SECTION 2.1. Sampling. The aquatic bryophyte, Fontinalis antipyretica Hedw., was collected at a permanent natural stream located in Serra de S~ao Mamede, Portalegre, Portugal. This species was selected because it is common and abundant in natural streams worldwide, it is easy to identify, and it has been used in previous studies regarding PAHs and metal biomonitoring.8,12 The samples were immediately transported to the laboratory in thermal boxes filled with the river water. At the laboratory, the moss was maintained in aerated commercial spring water of known composition for no more than one week at an average temperature of 15 °C. Moss thalli were then prepared for transplantation using nylon bags of approximately 15 15 cm as proposed by Cenci.13 Samples of approximately 100 g fresh weight (equivalent to dozens of individual thalli) were arranged inside the bag in a single layer to avoid superimposition of individual thalli. The moss bags (n = 12) were then transplanted to four different streams in the Lisbon municipality of Oeiras, a densely populated urban area (Figure 1). Previous works have shown that moss transplants don0 t significantly change their physiological performance when transplanted to the control site, namely regarding the ratio Fv/Fm of the chlorophyll a fluorescence.10 In this way, we assume that there is no significant physiological impact of the transplant during 3 months and thus no moss transplant was placed at the control stream. Moss transplants were exposed for 3 months from May to August 2008, after which they were collected into glass bottles and transported to the laboratory. The bottles were kept inside cold boxes during transport and in the laboratory. At the laboratory, each moss sample (control and transplants) was divided into three parts for PAH analyses, metal analyses, and chlorophyll a fluorescence measurements.
ARTICLE
2.2. Study Area. The study was conducted in the municipality of Oeiras, Portugal (Figure 1). Oeiras is a general urban area of 3748.8 inhabitants/km2 within the Lisbon metropolitan area. Lisbon metropolitan area, with a density of 1472.2 inhabitants/ km2, is the most densely populated area in Portugal, contrasting with the national average of 115.3 inhabitants/km2.14 Oeiras' main economic activities are in the tertiary sector (services): according to the data from the National Statistics Institute,14 80.8% of employees in Oeiras establishments work for the tertiary sector, while only 18.8% and 0.004% work for the secondary (manufacturing) and primary (involving the change of natural resources, or raw materials, into primary products) sectors, respectively. Like other urban areas in European countries, in this municipality urban areas (residential areas) can be found close to industries and to activities of the tertiary sector (Figure 1). 2.3. Land-Use Characterization. The geographic coordinates of each sampling site, as well as the land-use map for the study region provided by the municipality of Oeiras, were inserted into a geographic information system—ESRI ArcMap 9.3. Using this software, buffers (with radii from 100 to 2000 m) centered at each sampling site were drawn and the areas occupied by each land-use class that existed inside each buffer were computed. The land-use classes considered were industrial manufacturing areas (Industrial), urban or residential areas (Urban), facility areas or areas occupied by activities of the tertiary sector (Ter), and wooded areas (Trees). 2.4. PAH Analysis. All PAH analyses took place at the certified laboratory of the Portuguese Environmental Protection Agency (APA). Approximately 2 g of sample was extracted in a Soxhlet with 200 mL of acetonitrile for 24 h. After extraction, the extracts were concentrated by rotary vacuum evaporation and cleaned-up in a florisil column with 30 mL of acetonitrile as the eluting solvent. Then, the extracts were evaporated and concentrated with a gentle stream of purified N2 to 1 mL. The samples were analyzed by high-performance liquid chromatography (HewlettPackard), using two columns (Agilent C18 and Phenomenex C18), coupled to an ultraviolet fluorescence detector (FLD) and to an ultraviolet/visible detector (DAD/V-UV). The sixteen U.S. EPA-PAHs analyzed were acenaphtylene, naphthalene, fluorene, phenanthrene, fluoranthene, chrysene, benzo[a]anthracene, benzo[b]fluoranthene, benzo[k]fluoranthene, dibenzo[a,h]anthracene, benzo[g,h,i]perylene, acenaphtene, anthracene, pyrene, benzo[a]pyrene, and indeno[1,2,3-cd]pyrene. Concentrations below the detection limit were assumed to be 1/2 of the detection limit. PAH standards of Ultrascientific with an uncertainty of 5% were used. Recovery tests using Cryptogamic organisms were performed and showed percentage values between 60 ( 20 (for acenaphtlylene) and 107 ( 24 (for phenanthrene). 2.5. Metal Analysis. For metal analyses, 300 mg of control and transplant samples, dried at 50 °C for one week, were subjected to digestion in 4 mL of 67% nitric acid (HNO3). For each digestion, concentrations in acid blanks were subtracted from the transplant sample results. After the digestion, each solution (samples, blanks, and standards) was separated into three replicates and diluted with 10 mL of deionized water. Elemental zinc (Zn), copper (Cu), iron (Fe), chromium (Cr), manganese (Mn), and mercury (Hg) were analyzed by atomic absorption spectroscopy (Varian Techtron AA6, United Kingdom) using an air/acetylene chamber. Elemental lead (Pb), nickel (Ni), cadmium (Cd), and cobalt (Co) were analyzed by atomic absorption 3732
dx.doi.org/10.1021/es1036332 |Environ. Sci. Technol. 2011, 45, 3731–3738
Environmental Science & Technology
ARTICLE
Figure 1. Location of sampling sites in streams of the densely populated area of Oeiras, in Lisbon, Portugal, where moss transplants were exposed for 3 months (n = 12). One kilometer buffers are represented by circles. The study area faces Tagus River (one of the main rivers of the country) in the south band.
spectroscopy (CBC 932 plus) with a graphite chamber (GBC GF 3000). The analytical accuracy of the results was checked against the reference material referred by the Finish Forest Research Institute, Muhos Research Station.15 The results of the analyzed elements were within the confidence intervals of the certified values. 2.6. Measurements of Chlorophyll a Fluorescence. The effect of streamwater pollutants on the photosynthetic capacity of the moss transplants, after 3 months of exposure, was determined from fluorescence of chlorophyll a. A Mini Pam 101 Chlorophyll Fluorometer (Walz, Effeltrich, Germany) was used to measure chlorophyll a fluorescence in three samples of F. antipyretica from each sampling site. All samples were dark adapted for 10 min in saturated humidity before the fluorescence measurements. Dark adaptation maximizes oxidation of the primary quinine electron acceptor of PSII. After 10 min, the minimum fluorescence level with open PSII reaction centers (F0) was measured by a weak red measuring beam, followed by a saturation light pulse to determine the maximum fluorescence (Fm) level with closed PSII reaction centers. Variable fluorescence, Fv, is the difference between Fm and F0, and was calculated to obtain the parameter Fv/Fm. Chlorophyll fluorescence of control samples was also measured. It has been shown that Fv/ Fm parameter is a quantitative measure of the photochemical efficiency of photosystem II.12 The accumulation of excessive excitation energy can cause photoinhibition or photooxidation in the photosynthetic apparatus, and the reduced values of Fv/Fm indicate that a proportion of PSII reaction centers are damaged.12 2.7. Statistical Analysis. Pearson linear correlation coefficients between PAH or metal concentrations and chlorophyll fluorescence in the moss samples were calculated. Pearson linear
Figure 2. Enrichment factors (EF) of PAHs classified by ring number in the aquatic moss Fontinalis antipyretica, transplanted for 3 months to urban streams. The enrichment factors were obtained comparing the concentrations of PAHs prior to transplantation with the ones after the exposure period (n = 12).
correlation coefficients were also calculated for the correlation between PAH concentrations and the area covered by each class of the land use at each sampling site within buffers of increasing radii, from 100 to 2000 m (method adapted from ref 16). The coefficient (R-value) obtained for each correlation was plotted against the buffer radii. The land-use classes considered were Industrial, Urban, Tertiary, and Trees. A p-value 1.5) also occurred at both B4 and J2. In contrast, the sum of the 16 EPAPAHs decreased slightly in mosses transplanted to L3 and B2 in relation to control samples (Table 1). Because PAH profile differed among sampling sites within the same stream; most probably the pollutants have a more local origin and are quickly diluted by streamwater or deposited to sediments. This is very clear from the observation of PAHs in samples B2 and L3 which are in the middle of other sites that were shown to be more polluted downstream and upstream. As PAHs are lipophilic compounds, they tend to adhere to organic matter and suffer bioacumulation in biota. This will be further discussed in relation with the land use and distance of impact. Results also showed that the enrichment of 6-ring PAHs by mosses transplanted to stream A is particularly high, being 10fold more concentrated than the control site (Table 1). Mosses from site A1 were exceptionally enriched in the 2-ring PAHs (8fold more than controls) in contrast to all other transplanted 3734
dx.doi.org/10.1021/es1036332 |Environ. Sci. Technol. 2011, 45, 3731–3738
Environmental Science & Technology
ARTICLE
Figure 3. Pearson’s correlation coefficients (R) for the area occupied by each land-cover type (industrial, tertiary activities, urban (residential), and wooded areas-Trees) and the PAHs concentrations by ring number. The areas above the upper and below the lower dashed lines represent values with significant R values (p < 0.05).
mosses, which had enrichment factors between 0.19 and 2.20 (Table 1). The high value for 2-ring PAHs at site A1 is mainly due
to naphthalene, which suggests that an industrial or petrol station discharge might have occurred into this stream during the 3735
dx.doi.org/10.1021/es1036332 |Environ. Sci. Technol. 2011, 45, 3731–3738
Environmental Science & Technology
ARTICLE
Table 2. Pearson’s Coefficients for (1) Correlations between PAH and Metal Concentrations and (2) PAH Concentrations and Photosynthetic Capacity, Fv/FM, Measured in Aquatic Mosses Transplanted to Urban Streams for Three Months (n = 12)a
a
Significant correlations are shaded (p < 0.05).
exposure period. After a discharge into a stream, pollutants such as PAHs and the majority of metals tend to become diluted in the water and to accumulate in the sediments; therefore, water analyses usually provide values that are below detection limits and sediment analyses usually reflect long-term deposition of pollutants.19 For these reasons, illegal discharges are very difficult to observe and monitor using water because it reflects only shortterm pollution episodes or sediment analyses because they accumulate historical pollution episodes. The use of biological organisms that retain pollutants in their tissues might be a way of monitoring water pollution, integrating more time than a water analysis but less than a sediment one. Transplants are very flexible since we can use them for 3 months or more and obtain with this a rate of deposition since we are having in account the time dimension that is difficult to obtain with in situ organisms or sediment. Aquatic mosses tend to accumulate pollutants in their tissues during the exposure period and if any discharge is made during this period, mosses will probably reflect this contamination signal. Moreover, the accumulation of pollutants in aquatic macrophytes (as for example aquatic mosses) shows an evidence of PAH bioaccumulation in plant material that will feed other living organisms and contaminate the food-chain. The specific contamination of sampling site A1 supports our conclusion that contamination can be quite local and that, in as little as 3 months, mosses are able to integrate a profile of compounds to which they have been exposed. In terms of public health, the fact that some stream waters could be contaminated with such carcinogenic and mutagenic compounds is very important, because most water analyses are generally under detection limits for these kinds of contaminants and thus this methodology could work to detect localized water contamination. 3.2. PAHs in Relation to the Land Use and Metals. Tertiary sector (service) and industrial areas showed the highest correlation coefficients between land use and PAH concentrations (Figure 3). This is true for the sum of the 16 EPA-PAHs and
also for 2-, 3-, and 5-ring PAHs (Figure 3). Industrial areas had a significant effect if they were located less than 1000 m from streams (radii for which R-values were greater), whereas areas occupied by the tertiary sector had a significant effect if they were located less than 500 m from streams (Figure 3). To detect sources of PAHs in stream waters, correlations were calculated between PAH and metal concentrations in the moss transplants. Results are presented in Table 2. Sums of the 16 EPA-PAHs and the 2-, 3-, and 5-ring PAHs were significantly positively correlated with both Zn and Cu. This means that higher input of PAHs (especially of 2-, 3-, and 5-ring PAHs) is related to sources that also emit Zn and Cu. Urban runoff water quality problems are caused by the cumulative effects of many sources including heavy and light industry, road runoff and spills, and illegal dumping.20 The most significant sources of Zn urban runoff include atmospheric fallout, corrosion, tires, pavement water, automobile exhausts, exterior paint, road salt, and other terrestrial sources.21 The sources of Cu include corrosion of copper plumbing, electroplating waste, some algaecides, brake linings, and asphalt pavement wear.20 The U.S. EPA highlights 21 toxic substances that can mainly be assigned to road traffic; some heavy metals, such as Cu and Zn, are among them.21 It is important to remember that the main economic activities present in the study area are in the service sector; 80.8% of the employees in this area work for the tertiary sector, while only 18.8% and 0.004% work for the secondary and for the primary sectors, respectively.14 Associated with the tertiary sector is a set of roads with a considerable traffic flow, which might be responsible for the input of PAHs into the streams, as corroborated by the correlation between concentrations of the 16 EPAPAHs and concentrations of Cu and Zn in the transplanted mosses (Table 2). The 2-, 3-, and 5-ring PAHs also are positively correlated with these same elements (Table 2). As previously reported, urban runoff contains PAHs deposited on surfaces, as well as mobile related PAHs from gasoline and oil drips and 3736
dx.doi.org/10.1021/es1036332 |Environ. Sci. Technol. 2011, 45, 3731–3738
Environmental Science & Technology spills, exhaust products, tire particles, and bitumen from road surfaces.18 The main anthropogenic sources of PAHs in stream waters of our study area seem to be most likely related to traffic and urbanindustrial pollution. We suggest that enrichment of the HMW compounds in all transplanted mosses (Figure 2 and Table 1) reflects urban runoff, in accordance with other authors who demonstrated a relation between these PAHs and urban runoff and showed that the influence of traffic pollution appears to be reduced if the distance from a road is >500 m.20 3.3. Impact of PAHs on the Photosynthetic Capacity of Aquatic Mosses Transplanted to Urban Streams. Chlorophyll fluorescence measurements were performed to evaluate the impact of pollutants present on the stream waters in the photosynthetic capacity of the mosses after three months of exposure. The Fv/FM parameter measured in the transplanted mosses was generally low (between 0.06 and 0.57, which correspond to 8.6 and 80% of the control) compared to control samples (0.70). Some reduction of the Fv/FM parameter was expected, as the mosses were transplanted to a different place and thus had been subjected to different ecological conditions. The lack of significant linear Pearson’s correlations between moss PAH concentrations and the Fv/FM parameter in the same transplants indicates that the PAHs in the streams were not the most important pollutants causing damage to the photosynthetic apparatus of the aquatic moss transplants (Table 2). An exception was dibenzo [a,h]antracene, which had a significant negative correlation with Fv/FM, indicating that this compound could have influenced moss physiology (Table 2). Moss transplants are subjected to a wide range of pollutants and other nonfavorable ecological conditions in those urban stream waters, other than PAHs. This kind of measure is useful as it integrates the effect of all pollutants (known and unknown), including the synergistic and antagonistic effects and provides an estimate of pollution effects on plant biota. This study showed for the first time that urban streams seem to have a scattered contamination HMW-PAH; the areas occupied by tertiary (services) and industrial sectors have greater effects on moss transplants to urban streams, mainly for the sum of the 16 EPA-PAHs and for the 2-, 3-, and 5-PAHs, than other urban areas. These PAHs were correlated with sources that also emit Zn and Cu, which suggests a traffic-related origin. Industrial areas were most likely to be associated with enriched PAH concentrations in moss transplants if they were located
